摘要
阐述面向XGBoost的工控系统入侵检测分类模型,由于原始数据集存在数据样本不平衡问题,采用合成少数类过采样技术(SMOTE)对数据进行预处理,使用交叉验证方法寻找XGBoost最优参数,最后,在工控网络标准数据集上将本文算法与传统入侵检测方法进行对比实验。结果表明,对预处理后数据,基于XGBoost的工控入侵检测方法比传统方法具有更高的准确率。
This paper expounds the intrusion detection classification model of industrial control system for xgboost. Due to the imbalance of data samples in the original data set, the synthetic minority oversampling Technology(smote) is used to preprocess the data, and the cross verification method is used to find the optimal parameters of xgboost. Finally, the algorithm is compared with the traditional intrusion detection methods on the standard data set of industrial control network. The results show that for the preprocessed data, the industrial control intrusion detection method based on xgboost has higher accuracy than the traditional method.
作者
屠庆
鲍钰
TU Qing;BAO Yu(Shanghai Oriental Enviro-Industry Co.,Ltd.,Shanghai 200233,China;School of Software Engineering,East China Normal University,Shanghai 200062,China)
出处
《集成电路应用》
2022年第2期52-55,共4页
Application of IC
